ABSTRACT
Consulting statisticians do not always get the chance to design the experiments for which they must help construct appropriate analyses. Instead, the statistician must rst identify the type of designed experiment the researcher has employed. The rst and most important step in the identi cation process is to determine if more than one size of experimental unit has been used, and if so, to identify each size of experimental unit. As will become evident in this section, each size of experimental unit will have an associated design structure and treatment structure. After the different sizes of the experimental units have been identi ed, the model for carrying out an appropriate analysis can be constructed by combining the models used to describe the design structure and treatment structure corresponding to each size of experimental unit.
